N-able genAI evangelist: As ROI stagnates, MSPs have an opportunity to become 'trusted advisors'

Nicole Reineke, distinguished product manager for AI strategy, talks genAI use cases, ROI and service opportunities

MSPs continue to evaluate the use cases and service opportunities opened up by generative AI implementations.

But several consultancies have suggested in recent weeks that the return on investment for end-users might be years in the making and may not materialise until mid-2025.

A Gartner study from May found that the biggest barrier to end-user adoption of the tech is in demonstrating its value.

Meanwhile, multiple partners have shared with CRN that they are in exploration mode on the use cases of the technology, even as the likes of Bytes, Performanta, Node4 and Crayon go big on genAI solutions, notably Copilot.

Meanwhile, IT management vendor N-able's recently published Horizon report found that 25 per cent of MSPs are currently not using generative AI at all - potentially missing an opportunity to optimise their services.

As the tech continues to evolve and the leaders emerge, forward-thinking MSPs are exploring ways to harness its potential while navigating the complexities that come with adoption.

Nicole Reineke, distinguished product manager for AI strategy at N-able (pictured), sat down with CRN to shed light on the obstacles and prospects that lie ahead for MSPs embracing generative AI.

One of the primary hurdles is the sheer complexity of the technology itself.

"Explainability is a huge issue with generative AI," Reineke says.

"There's a really big need for transparency and explainability in any generative AI application.

"This lack of transparency can raise concerns, particularly in regulated industries where accountability and compliance are paramount.

"MSPs must carefully consider the ethical implications and ensure they have robust governance frameworks in place."

Despite the challenges, generative AI presents a wealth of opportunities for MSPs to differentiate their services and drive innovation.

Reineke highlights security as a significant area of potential.

"Anything that we can do around security is a huge opportunity for improvement, and it's an area of high risk for MSSP customers," she says.

"As we get better at generative AI, and we know more information about each of those end devices, the MSPs know more information about each of those end devices, there's an opportunity in proactively identifying and saying ‘Hey, there may be something here that we should do something about. You've drifted away from your security protocol.' And then taking action to remedy it."

Automation across customer communications

Another promising avenue is the automation of routine tasks, freeing up valuable resources for more strategic initiatives.

"We really want to only use generative AI to augment humans," Reineke clarifies.

"We do not want to use it to automate humans."

The customer support use case has emerged at an early success for MSPs because of the measurable nature of the function, she says.

"Because we're not reducing the number of machines that are being bought and we're not reducing some of the capital expenditures, we actually have to understand what the current human costs are related to any activity that you're starting to augment," Reineke explains.

"One of the reasons the customer support use case has become so widely adopted is because we can actually measure, for example, this person closes 22 tickets every two hours, and bringing in AI, we can deflect five of those tickets. And so we can do a very clear mathematical calculation."

These kinds of measurable applications go some way towards easing the ROI maths, according to Reineke.

"You have to understand what your current costs are for the activities that you're looking to augment with AI," she says. "Customer support and marketing are areas where AI adoption has been easy because the outcomes are measurable."

From there, the business case for investment becomes a lot easier to articulate – and communicate to clients.

Skills gaps may hold the key for MSPs

Another significant barrier to mass adoption has been the AI skills gap and this is particularly true when it comes to data engineering.

Reineke estimates that data engineering comprises 75-80 per cent of AI project efforts, with companies needing support to properly curate and label training data.

This presents another significant opportunity for MSP firms, particularly those with in-house data engineering expertise, to build out AI-adjacent services lines.

As the generative AI landscape continues to evolve, MSPs have a unique opportunity to position themselves as trusted advisors, guiding clients through the complexities of adoption while offering secure, isolated environments for running AI models and applications.